Session Recap: State-Provided Paid Leave
November 12, 2014 09:00 AM
By Chiho Song, University of Washington
Socially Insuring Family Leave: The Relationship Between Public Policy, Paid Family Leave, and Economic Well-Being by Linda Houser, Widener University; Thomas Vartanian and Jenifer Norton, Bryn Mawr College
Maternity Leave Job Quitting By Less-Educated Mothers: What is the Role of State-Level Safety Net and Maternity Leave Policies? by Marci Ybarra, University of Chicago; Heather Hill, University of Washington
TANF Generosity, State-Provided Maternity Leave and the Material Wellbeing of Low Income Families with Infants by Marci Ybarra and Alexandra Stanczyk, University of Chicago; Yoonsook Ha, Boston University
Discussant: Lawrence Berger, University of Wisconsin–Madison
Houser and Vartanian examined the impact of the maternity support policies, referred to as “Temporary Disability Insurance (TDI)” or Paid Family Leave (PFL) policies on wellbeing of low-income families. To this aim, multivariate regression analyses (linear/logistic) and a difference-in-difference (DID) model are used based on the National Longitudinal Survey of Youth (from 1997 to 2009) Panel. The analytic samples are men and women who reported a birth with having worked at least twenty hours per week in the three or four months prior to the birth. In addition, for controlling the individual-level characteristics and state-level policy variation, related factors such as demographics, SES, and state dummies are included in their model.
There are two main outcomes: 1) Likelihood of using paid family leave and 2) Duration of using paid family leave. The effects of the maternity support policies on the outcomes are evaluated by gender and by presence/absence of maternity support (TDI or PFL) policies. Their main findings are 1) the likelihood of taking paid leave following the birth of a child for women within states with maternity support policies are twice higher than those within states without, 2) the effect of maternity support on the likelihood of taking paid leave is stronger for low-income women, 3) women in California (having a long tradition of maternity support policies compared to other states) are more likely to take paid leave than those in other states, and 4) Comparing pre- and post-PFL period (1997-2004 vs. 2005-2009), new fathers in California are more likely to use paid leave which is averagely more than twice higher than before. Berger discussed the papers limitations that should be overcome for future research, including 1) selection bias, 2) limited age range (ages<=30), 3) bias in survey questions (meaning/interpretation of “leave” question, 4) current financing problems, and 5) awareness and uptake.
Ybarra next presented her work with Hill, sharing her investigation of the role of safety net and maternity leave policies on maternity-leave job quitting for less-educated mothers. Their basic concern is that there is a gap in knowledge regarding the relation between the safety net and maternity-leave job quitting. To achieve this aim, the five multinomial regression models are used based on the Survey of Income and Program Participation (SIPP; stacked 2001, 2004, and 2008 fertility modules).
Ybarra and Hill found three primary outcomes: job quitting during pregnancy, timing of job quitting (only if females quitted her job during pregnancy), and job quitting post-birth. Their hypothesis is that females living in states with more generous safety net programs in conjunction with less generous maternity leave policies are more likely to quit her job as a form of maternity leave. This preliminary evidence does not indicate strong relation between safety net generosity and maternity-leave among less educated mothers. Some points to be refined for future research that Berger discussed is 1) Difference-in-difference (DID) technique is required for estimating the effects of state-level safety net policy changes (within and across states) on maternity-leave job quitting and 2) Sub-group analysis of states with unemployment rates above the national average is needed.
Stanczyk scrutinized how TANF and/or state-provided paid leave programs impact the material wellbeing of low-income families in the period surrounding birth. To this aim, this study attempts to predict TANF use and material hardship (i.e. difficulty paying expenses or utilities, and rent/mortgage) assuming that these two outcomes may be affected by the state-level role of TANF generosity (measured by time limits, length of work exemptions for new mothers, eligibility of pregnant women, diversion programs, and average benefit level) and Paid Maternity Leave provisions in the period surround birth.
The authors' study is based on the pooled panel of the Survey of Income and Program Participation (SIPP) that is known as nationally representative sample with monthly TANF participation records and related measures of material hardship. Based on this data, state-year policy variables derived from the Urban Institute’s Welfare Rules Database are also attached to the analytic data. The sample is mothers who are below 200% Federal Poverty Line with having sufficient observations for pre- and post-birth. They use multivariate analyses (probit models) to predict post-birth TANF participation and material hardship.
Stanczyk discussed their major findings, which were that post-birth TANF participation is more likely among less advantaged mothers; higher TANF benefits and earnings allowances increase the likelihood of TANF use; the relation between taking paid leave and post-birth TANF use is insignificant after controlling for state TANF generosity and individual/household characteristics; post-birth material hardship is more likely among less advantaged mothers, and; that the association between taking paid leave and post-birth material hardship is significant, but there is an inconsistency in its direction.
© 2020 Association for Public Policy Analysis & Management. All Rights Reserved.